Heya folks.

Finally having dealt with (other than a new oil return seal) the dead turbo (Put in a slightly bigger td04hl-15g) on my 740, I'm back at the problem I was having before it died.

When running above idle speeds, the injectors will stop pulsing for anywhere from an instant for up to 10 seconds. I determined this by backprobing an LED into the injector wiring and seeing the LED stop flashing every time the engine cut out. Sparks continued uninterrupted.
Signal to the fuel ECU from the ICU looks consistent.

Fuel ECU is throwing codes about the AMM, TPS, and lean/rich conditions. The AMM and TPS feed from a common circuit powered by the main fuel relay - I've resoldered 2 relays and tried both of them, but it persists with both.

I pulled the fuse panel apart and found #1 fuse a bit melted. Took off the top cover and wire-brushed all of the connectors and reassembled - no change.

I'm about to go borrow a -560 (problematic w/fuel circuit failure) ECU and see if it works.


I was just doing some more research, and came across a mention of a traction control system called Luxor that shuts off injectors when wheel spin is detected, and also that the injectors shut off when the throttle switch says the throttle is closed "when coasting". I wonder if the throttle switch or its wiring is shorting out.

I'm interested in exactly how the wiring runs and how the injectors are grounded by the ECU. I think it's pin 18 responsible for grounding them, the RSR feeds them power, the main fuel relay feeds the RSR, and the ECU grounds/energizes the main fuel relay to turn it on.

Is there a logic tree somewhere of how the injectors pulse? Can someone help me build one?

Cheers!

A-HA! I FIGURED THE DARNED THING OUT!

Thanks to numerous suggestions here, I checked the voltage between the TPS and ground. It was inconsistent to say the least. I used a cheapo analog meter to watch the voltage, and every time it dropped far enough, the injectors stopped.

I cut open the harness and replaced both orange wires (spliced together at the back side of the #1 (I think) pin. Problem gone. I am so thrilled. Thanks everyone for the help!

For a deeper description of the problem, looking at a wiring diagram shows that there is one wire from the ECU (LH2.4) to the pin on the TPS and one from the ICU (the EZK unit). When the idle switch is open, the ICU sends +12V to the ECU. When the idle switch closes, the voltage drops off as the center TPS pin (ground) is a path of less resistance.
To check this, first I backprobed the TPS connector and watched the voltage. Then I disconnected the ECU, ICU, and TPS, and checked for resistance to ground. Then I disconnected the harness connectors at the strut towers and checked for resistance to ground.

It was probably only necessary to replace one, but I wasn't sure which one was to blame.

Cheers! I should be back on the forum more often now.

PS As for the other two codes (AMM and mixture errors), a bad relay and a relay contact pushed out of its socket were to blame.

Started stalling on me on the highway on the way home last night.

I finally traced the problem down - I didn't use the harness guide clamps when I did my engine replacement because I didn't know where they went. The injector and TPS wiring was melting through and shorting out on the EGR valve.

All fixed now, I think.

glad you caught it.

however, never use an analog meter on any car when checking voltage or ohms and an ecu. analogs do not have enough internal resistance and will allow voltage to pass. digital voms have 10k resistance built in to prevent this. good luck, chuck.

Hello,

Good to know the news! Rejoice!

I think its the tower harness connectors were the culprit. Over time the insulators have disintegrated exposing those contacts to the elements.

I have soldered all these wires together (dispensed all connectors) and heat-shrinked them. That should take them out of the equation. And because these wires don't move much (much less vibrate) I think soldering than crimping is good enough.

Check also there's a square rubber gasket within the TPS connector. Its to secure the connection (less vibration) and seal them from the elements (and prevent future issues).

I think its all about connection integrity that may have been the issue (well known to affect 740s).

One more thing about the ECU codes. They are cumulative. Old issues detected are not automatically erased once it is fixed. So its helpful to clear the codes before troubleshooting.

Regards,
Amarin.

you first need to figure what is quitting-power or ground.

if it's power, then it's the relays you mentioned. there's also a ground behind the lh headlight for the rs relay. had that 1 stall a car when the high beams came on.

has the head been off? had a 740t that the dealer pulled the head and trapped the wire from the relay to the rs relay and would cut power off on hard accel. replace the wire and all was better.

if it's ground, the it's the ecu inputs missing or the ecu itself. just replaced an ecu on an lh 2.2 for that. pin 13 would quit when all else was good.

make sure the throttle sw is ok, it will command injectors off when it's closed at idle.

the melted fuse concerns me, never seen that before. shouldn't be that high an amperage draw. good luck, chuck.

The head hasn't been off since I put this engine in a few years ago. It might have been before then, but there were no wires trapped.

I'm wondering if maybe the TPS is flaking out, but it seems like it's the ECU. It's been throwing TPS, AMM, and lean/rich codes.

Hmmm. How to see if the ground is disappearing or the power. Well, power = one wire into the injector harness, the other to ground - if the light stops flashing with the cut-outs, power. If I backprobe with a multimeter and check for resistance to ground, it should skyrocket when the engine cuts out if it's losing ground.

I've cleaned all the ground points I can find - manifold, firewall, both headlights, and passenger footwell.


The fuse was warped, the socket was a little warped, but not too bad. Apparently, it's not unusual.

i used 2 dvoms to catch the ecu i mentioned. when the ground side showed 0 and the power said 12v, i had it.

the excess codes reminded me---i had an lh2.4 car that threw lotsa codes, some repeated, some didn't. i had a tester that i let the cust drive the car on and it all went away. new ecu fixed it. still have that test ecu, served me well. not my memory, though. good luck, chuck.

Hello,

"When running above idle speeds, the injectors will stop pulsing for anywhere from an instant for up to 10 seconds. I determined this by backprobing an LED into the injector wiring and seeing the LED stop flashing every time the engine cut out"

I'm confused by the above statement. If the injectors have stop pulsing for 10 seconds meaning no fuel injected into the engine for 10 seconds, the engine must have surely died. No engine can stay running after 10 seconds of not any injector firing fuel. Unless of course the car is on the move.

Are you rigging this test in a moving car? Of course there will be fuel cut out when slowing down or running down hill. And the ECU knows this by feedback from the TPS, CPS, AFM and rear axle sensors. So it cuts fuel. Otherwise the engine RPM wouldn't go down. Otherwise fuel will be wasted during running down hill.

I think you may have placed undue interpretation on this 'noid light' LED test. It's supposed to be a simple test where the LED is connected to the injector electrical socket to indicate whether the injector is receiving electrical impulse or not (i.e firing or not). Its for 'no start' condition where this 'noid light' test would point the problem elsewhere e.g. faulty ignition system or fuel pump. I don't think this test is appropriate to map the injector running pattern.

If your engine is running well at idle but cuts out outside idle (even when throttle is held steady) then something else is missing. I think even slight reduction in throttle would reduce the airflow which reduces engine RPM. This split second reduction in rotation is sensed by the CPS (via the 60 plus 'holes' in the tone ring on flywheel) and ECU cuts the fuel. This event is also 'seen' by the AFM which measures the change in the platinum wire's resistance due to reduced airflow. Maybe this is what you've seen by those momentary 'missing pulses' from the LED. I may be wrong as you know how steady you've held the throttle.

Anyways I've read that in Porsche's 944 the injectors are wired to two grounding terminals at the ECU (but these terminals are connected together inside). This is to reduce the electrical load on the terminals. Our ECU have only one grounding terminal (that's pin18) for all the injectors. Check its connection.

Load test the ECU terminal electrical connector (for pin18) by inserting something that's 0.04 inch wide (a needle maybe). It should grip the item strongly. I'm not sure whether this applies to our cars but its a free check. You could have a loose connector. Load test also other connectors in the ECU electrical plug. Injector plugs too.

Regards,
Amarin.